The PDFs on this page have been archived. Links will take you to documents on the National Archive Website.

Sir John Bourn, head of the National Audit Office, reported to Parliament today that the failure to introduce the Benefits Payment Card had cost the Department of Social Security some £127 million in nugatory system development. Delays in the Card programme, which was part of a pioneering Private Finance Initiative (PFI) project designed also to automate post offices, led to a radical restructuring of the project in May 1999. Over £300 million has been forgone in lost opportunities to prevent benefit fraud.

Jump to downloads

The other purchaser, Post Office Counters Ltd, and the supplier ICL made charges in their accounts for £571 million and £180 million respectively. The NAO report highlights important lessons for government and the wider IT community on handling major projects, particularly on how to manage risks.

In 1996 the Benefits Agency and Post Office Counters Ltd jointly awarded a Private Finance Initiative PFI contract to Pathway, a subsidiary of the ICL computer services group. The Benefits Payment Card project was intended to replace by 1999 existing paper based methods of paying social security benefits with a magnetic stripe payment card, and to automate the national network of post offices through which most benefits are paid. The project was vast and complex, and estimated to cost some £1 billion in payments to Pathway. It was also one of the first PFI Information Technology contracts. The project was a tripartite venture, requiring all three parties involved to meet their contracted obligations for it to be successful.

The purchasers used Private Finance because they did not expect to have the capital resources to develop the Benefits Payment Card themselves and because PFI was the preferred option for all new IT projects at that time. It would also transfer to the private sector risks of developing and delivering a working system and preventing fraud. The key expected benefit for the Department lay in saving over £100 million a year by preventing the fraudulent use of order books and girocheques. This meant that delay in delivery would erode the viability of the project to them.

By October 1996 the contracting parties had successfully implemented a limited version of the system. But designing and developing a fully functional system proved much more complex and took much longer than had been expected. The programme at the time the contract was signed assumed that it would take ten months to start a live trial of the full system. In fact, although partial trials had started this stage had not been reached at the time the contract was terminated nearly three years later.

Following extensive review and negotiation, in May 1999 the government decided that removing the payment card from the project offered better value for money than complete cancellation. Very importantly, it would simplify the project, reduce the risks and protect the early automation of the Post Office by 2001, and was preferable to continuation.

"There may be a temptation to think that the Payment Card project failed solely because it was large and complex or because it was a pioneer PFI project. This was not the case. Various factors contributed to the project’s failure and their effects are difficult to disentangle. Important reasons were divided control, insufficient time for specifying the requirement and piloting, and that a shared, open approach to risk management was not achieved."

Our examination found that:

The report highlights key lessons on risk management, the procurement of complex IT systems, and procurement by more than one purchaser, including:

In March 2000 the Treasury PFI taskforce issued new guidance on PFI contracts for Information Technology. In May 2000 the Cabinet Office published recommendations for improving the way in which the government approaches and manages major Information Technology projects. Had this guidance existed and been applied in the case of this project, it may have substantially reduced the risk of it failing, or alternatively have led to the project not proceeding in the way it did.

Sir John Bourn


Publication details

Latest reports